An etiologic prediction model incorporating biomarkers to predict the bladder cancer risk associated with occupational exposure to aromatic amines: a pilot study

Table 3 Classification of 15 AA-related BC cases (D) and 184 other BC cases (−D) according to the full model (left panel) or short model (right panel) of logistic regression

Classified	True		Total	Classified	True		Total
Classified	D	–D	Total	Classified	D	–D	Total
Positive	13	15	28	Positive	13	47	60
Negative	2	169	171	Negative	2	137	139
Total	15	184	199	Total	15	184	199
Sensitivity			86.7%	Sensitivity			86.7%
Specificity			91.9%	Specificity			74.5%
Positive predictive value			46.4%	Positive predictive value			21.7%
Negative predictive value			98.8%	Negative predictive value			98.6%
Correctly classified			91.5%	Correctly classified			75.4%
Net Benefit per 100 patients			+4.9	Net Benefit per 100 patients			+ 4.1

The risk threshold R (optimal cut-off point of predicted probability provided by the program) was 0.181 and 0.093 in the left and right panel, respectively. Net Benefit per 100 patients calculated from the above values

ISSN: 1745-6673